
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import matplotlib.pyplot as plt


import os
for dirname, _, filenames in os.walk('/kaggle/input'):
    for filename in filenames:
        print(os.path.join(dirname, filename))


import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots


df = pd.read_csv("C:/Users/kumar/OneDrive/Documents/IV_J_COMPONENT/Space_Corrected.csv")
df.head()


df.columns

Index(['Unnamed: 0', 'Unnamed: 0.1', 'Company Name', 'Location', 'Datum',
       'Detail', 'Status Rocket', ' Rocket', 'Status Mission'],
      dtype='object')


df = df.drop(["Unnamed: 0", "Unnamed: 0.1"], axis = 1)
df.head()


df.describe()


df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4324 entries, 0 to 4323
Data columns (total 7 columns):
 #   Column          Non-Null Count  Dtype 
---  ------          --------------  ----- 
 0   Company Name    4324 non-null   object
 1   Location        4324 non-null   object
 2   Datum           4324 non-null   object
 3   Detail          4324 non-null   object
 4   Status Rocket   4324 non-null   object
 5    Rocket         964 non-null    object
 6   Status Mission  4324 non-null   object
dtypes: object(7)
memory usage: 236.6+ KB


ds = df["Status Rocket"].value_counts().reset_index()
ds


fig = px.pie(ds, values = "Status Rocket", names = "index", title = "Rocket Status")
fig.show()


ds = df["Status Mission"].value_counts().reset_index()[:3]
ds


fig = px.bar(ds, x = "index", y = "Status Mission", title = "Mission Status")
fig.show()


sum(pd.isna(df.loc[:, " Rocket"]))

3360


df_ = df.dropna(subset = [" Rocket"], axis = "rows")
len(df_)

964


sum(pd.isna(df_.loc[:, " Rocket"]))

0


df_.loc[:, " Rocket"]

0        50.0 
1       29.75 
3        65.0 
4       145.0 
5       64.68 
         ...  
3855     59.0 
3971    63.23 
3993    63.23 
4000    63.23 
4020    63.23 
Name:  Rocket, Length: 964, dtype: object


df_.loc[:, " Rocket"] = df_.loc[:, " Rocket"].fillna(0.0).str.replace(",", "")
df_.loc[:, " Rocket"] = df_.loc[:, " Rocket"].astype(np.float64).fillna(0.0)

C:\Users\kumar\anaconda3\lib\site-packages\pandas\core\indexing.py:1843: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


df_d = df_[df_.loc[:, " Rocket"] < 1000]
plt.figure(figsize = (22, 6))
sns.histplot(data = df_d, x = " Rocket", hue = "Status Rocket")
plt.show()


df.head()


df_money = df_.groupby(["Company Name"])[" Rocket"].sum().reset_index()
df_money = df_money[df_money[" Rocket"] > 0]
df_money.head()


df_money_ = df_money.sort_values(by = [" Rocket"], ascending = False)[:15]
df_money_.head()


fig = px.bar(df_money_, x = "Company Name", y = " Rocket", title = "Total Spent Money for Each Company")
fig.show()

	Unnamed: 0	Unnamed: 0.1	Company Name	Location	Datum	Detail	Status Rocket	Rocket	Status Mission
0	0	0	SpaceX	LC-39A, Kennedy Space Center, Florida, USA	Fri Aug 07, 2020 05:12 UTC	Falcon 9 Block 5 \| Starlink V1 L9 & BlackSky	StatusActive	50.0	Success
1	1	1	CASC	Site 9401 (SLS-2), Jiuquan Satellite Launch Ce...	Thu Aug 06, 2020 04:01 UTC	Long March 2D \| Gaofen-9 04 & Q-SAT	StatusActive	29.75	Success
2	2	2	SpaceX	Pad A, Boca Chica, Texas, USA	Tue Aug 04, 2020 23:57 UTC	Starship Prototype \| 150 Meter Hop	StatusActive	NaN	Success
3	3	3	Roscosmos	Site 200/39, Baikonur Cosmodrome, Kazakhstan	Thu Jul 30, 2020 21:25 UTC	Proton-M/Briz-M \| Ekspress-80 & Ekspress-103	StatusActive	65.0	Success
4	4	4	ULA	SLC-41, Cape Canaveral AFS, Florida, USA	Thu Jul 30, 2020 11:50 UTC	Atlas V 541 \| Perseverance	StatusActive	145.0	Success

	Company Name	Location	Datum	Detail	Status Rocket	Rocket	Status Mission
0	SpaceX	LC-39A, Kennedy Space Center, Florida, USA	Fri Aug 07, 2020 05:12 UTC	Falcon 9 Block 5 \| Starlink V1 L9 & BlackSky	StatusActive	50.0	Success
1	CASC	Site 9401 (SLS-2), Jiuquan Satellite Launch Ce...	Thu Aug 06, 2020 04:01 UTC	Long March 2D \| Gaofen-9 04 & Q-SAT	StatusActive	29.75	Success
2	SpaceX	Pad A, Boca Chica, Texas, USA	Tue Aug 04, 2020 23:57 UTC	Starship Prototype \| 150 Meter Hop	StatusActive	NaN	Success
3	Roscosmos	Site 200/39, Baikonur Cosmodrome, Kazakhstan	Thu Jul 30, 2020 21:25 UTC	Proton-M/Briz-M \| Ekspress-80 & Ekspress-103	StatusActive	65.0	Success
4	ULA	SLC-41, Cape Canaveral AFS, Florida, USA	Thu Jul 30, 2020 11:50 UTC	Atlas V 541 \| Perseverance	StatusActive	145.0	Success

	Company Name	Location	Datum	Detail	Status Rocket	Rocket	Status Mission
count	4324	4324	4324	4324	4324	964	4324
unique	56	137	4319	4278	2	56	4
top	RVSN USSR	Site 31/6, Baikonur Cosmodrome, Kazakhstan	Tue Jun 26, 1973	Cosmos-3MRB (65MRB) \| BOR-5 Shuttle	StatusRetired	450.0	Success
freq	1777	235	2	6	3534	136	3879

	index	Status Rocket
0	StatusRetired	3534
1	StatusActive	790

	index	Status Mission
0	Success	3879
1	Failure	339
2	Partial Failure	102

	Company Name	Rocket
0	Arianespace	16345.00
1	Boeing	1241.00
2	CASC	6340.26
3	EER	20.00
4	ESA	37.00

	Company Name	Rocket
14	NASA	76280.00
0	Arianespace	16345.00
21	ULA	14798.00
16	RVSN USSR	10000.00
2	CASC	6340.26